J. Lat. Amer. Stud. , – Printed in the United Kingdom #  Cambridge University Press  An ‘Organised Disorganisation’: Informal Organisation and the Persistence of Local Party Structures in Argentine Peronism*

STEVEN LEVITSKY

Abstract. This article attempts to fill the void in research on the Justicialista Party (PJ) organisation. Challenging accounts of the contemporary PJ as a weak, personalistic organisation, it argues that the party maintains a powerful base-level infrastructure with deep roots in working- and lower-class society. This organisation has been understated by scholars because, unlike prototypical working class party structures, it is informal and highly decentralised. The PJ organisation consists of a range of informal networks – based on unions, clubs, NGOs and activists’ homes that are largely unconnected to the party bureaucracy. These organisations provided the government of Carlos Menem with a range of benefits in the s, particularly in the realm of local problem-solving and patronage distribution. Yet they also constrained the Menem leadership, limiting its capacity to impose candidates and strategies on lower-level party branches.

The Argentine Justicialista – or Peronist – Party (PJ) has long posed a puzzle for analysts. Although Peronism’s electoral strength is beyond dispute, the weakness and inactivity of the party bureaucracy and formal leadership bodies have led scholars to describe its organisation as virtually " non-existent. Scholars dismiss the original Peronist party as a ‘cadaver’ # or ‘little more than an appendage of state bodies’. Similarly, the $ contemporary PJ has been described as a ‘mere nameplate’ or an

Steven Levitsky is Assistant Professor of Government at Harvard University. * Research for this article was made possible by grants from Fulbright-Hays and the Social Science Research Council. I am grateful to Javier Auyero, Frances Hagopian, David Samuels and three anonymous reviewers for their helpful comments on earlier drafts. " Fe! lix Luna, PeroT n y su tiempo (Buenos Aires, ), p. . Luna writes that ‘no one will be able to write the history of the Peronist party between  and  because it never existed’. # Marcelo Cavarozzi, Peronismo y radicalismo: transiciones y perspectivas (Buenos Aires, ), p. . For a critique of this view, see Moira Mackinnon, ‘Sobre los orı!genes del partido peronista: Notas introductorias’, in W. Ansaldi, A. Pucciarelli, and J. Villarreal, Representaciones inconclusas: Las clases, los actores, y los discursos de la memoria, ‰‚–‰„† (Buenos Aires, ). $ Gabriella Cerruti, El Jefe: Vida y obra de Carlos SauT l Menem (Buenos Aires, ), p. .  Steven Levitsky ‘electoral committee’ run by a small circle of ‘operators’ in Buenos % Aires. Another look at the PJ organisation, however, reveals a strikingly different picture. The contemporary PJ maintains a massive base-level infrastructure, and its membership of nearly four million makes it one of the largest democratic parties in the world. Moreover, Peronism’s deep social and organisational roots in working and lower class society have enabled it to survive decades of proscription, the death of its charismatic founder, and, more recently, the repudiation of its traditional socio- economic programme. How can the PJ simultaneously be so weak and so strong? A major reason for this confusion is that when scholars look for the PJ, they tend to look in the wrong place. Attention to the weakness of the PJ’s formal structure tends to obscure the vast informal organisation that & surrounds it. The Peronist organisation consists of a dense collection of personal networks – operating out of unions, clubs, non-governmental organisations, and often activists’ homes – that are largely unconnected to (and autonomous from) the party bureaucracy. Although these networks are not found in the party’s statutes or records, they provide the PJ with extensive linkages to working and lower class society. Notwithstanding the existence of an impressive body of literature on Peronism, little ' research has been done on the PJ organisation, and virtually no work has been done on the party’s informal structure. As a result, we know little about how the PJ functions, particularly at the local level. This article takes a step toward filling that gap. Drawing on data collected during an extensive study of local Peronist organisations in Greater Buenos Aires, the article examines how the PJ is organised and functions internally. The article is divided into two sections. The first outlines the origins and contemporary structure of the Peronist party organisation. Challenging characterisations of the party as a weak, personalistic organisation, it argues that the PJ is best understood as an informal mass party. Peronism maintains a powerful base-level infra-

% Marcos Novaro, ‘Menemismo y peronismo: Viejo y nuevo populismo’, in Ricardo Sidicaro and Jorge Mayer (eds.), PolıTtica y sociedad en los ang os de Menem (Buenos Aires, ), p. ; and Vicente Palermo and Marcos Novaro, PolıTtica y poder en el gobierno de Menem (Buenos Aires, ), pp. –. & This focus on informal organisation was inspired by Guillermo O’Donnell’s recent discussion of informal institutions in democratic regimes. See O’Donnell, ‘Illusions About Consolidation’, Journal of Democracy, vol. , no.  (), pp. –. ' Exceptions include Vicente Palermo, Democracia interna en los partidos (Buenos Aires, ) and Ana Marı!a Mustapic, ‘El Partido Justicialista: Perspectiva histo! rica sobre el desarrollo del partido. La estructura del partido’, unpublished manuscript, Torcuato Di Tella University, Buenos Aires, . Local Party Structures in Argentine Peronism  structure, a large activist base and extensive linkages to working and lower class society, but unlike prototypical mass parties, its organisation is informal and highly decentralised. The second section examines the relationship between the PJ and Carlos Menem in the s. Contrary to many conventional accounts, it argues that Menem’s relationship to the PJ rank-and-file was always mediated by powerful local organisations. These organisations provided the Menem government with a range of political benefits, including vast human resources and extensive channels for policy implementation, patronage distribution, and local problem-solving. Yet they also constrained the Menem leadership, limiting its capacity to impose candidates and strategies on lower-level branches. Indeed, local branches routinely rejected – or ignored – instructions from the national leadership, following strategies that had little to do with either Menem or his neo-liberal programme. The article’s focus on the PJ’s internal structure serves two broader analytic goals. First, it highlights the importance of studying informal patterns of party organisation. In many Latin American parties, the organisations that exist ‘on the ground’ differ substantially from those outlined in the statutes, and formal rules and procedures are fluid, manipulated, and even ignored. Nevertheless, relatively little effort has been made to investigate, conceptualise, or theorise these organisational forms. Research on the informal structures of Latin American parties can play an important role in broadening and refining the existing literature on parties and party organisations. Because this literature is based almost ( entirely on studies of advanced industrialised countries, where parties tend to be relatively well-institutionalised, it incorporates little variation on these dimensions. This failure to incorporate a wider range of cases has ) important theoretical costs. Parties with informal structures and fluid internal rules function differently – and with different consequences – * from well-institutionalised or bureaucratic parties. In order to be able to assess these differences systematically we need a more thorough understanding of how Latin American parties work. This article takes a

( For example, Maurice Duverger, Political Parties: Their Organisation and Activity in the Modern State (New York, \); Angelo Panebianco, Political Parties: Organisation and Power (Cambridge, ); Herbert Kitschelt, The Transformation of European Social Democracy (New York, ), Richard S. Katz and Peter Mair, How Parties Organize: Change and Adaptation in Party Organisations in Western Democracies (London, ). ) Scott Mainwaring makes a similar point in his recent work on the Brazilian party system (Mainwaring, Rethinking Party Systems in the Third Wave of Democratisation: The Case of Brazil (Stanford, ), pp. –). * See Steven Levitsky, Crisis, Party Adaptation, and Regime Stability in Argentina’, Party Politics, vol. , no.  (), pp. –; Mainwaring, Rethinking Party Systems in the Third Wave of Democratisation.  Steven Levitsky step in that direction, going beyond party statutes and formal organisations to examine the informal rules of the game that structure the internal life of the PJ. The article also seeks to contribute to the emerging literature on the politics of economic reform in Argentina. Early analyses of the Menem period characterised the reforms as a kind of neoliberal revolution from above, imposed by a powerful president acting at the margins of the PJ, major interest groups, and the legislature and other democratic "! institutions. Indeed, the Menem presidency was characterised – along with those of Fernando Collor and Alberto Fujimori – as a near- "" archetypal case of ‘neo-populism’. Recent research suggests that such accounts may have been somewhat overstated. For example, scholars have shown that Menem’s capacity to impose reforms unilaterally was in fact relatively limited, and that the passage of most major reforms required substantial concessions to governors, business and labour leaders, PJ "# legislators, and other key social and political actors. Other scholars have challenged the notion that the PJ was marginalised under Menem, arguing that the party was critical to Menem’s success in both the "$ legislative and electoral arenas. The data presented here provides further

"! Guillermo O’Donnell, ‘Delegative Democracy’, Journal of Democracy, vol. , no.  (), pp. –; Christopher Larkins, ‘The Judiciary and Delegative Democracy in Argentina’, Comparative Politics, vol. , no.  (), pp. –; Larry Diamond, Developing Democracy: Toward Consolidation (Baltimore, ), pp. –. "" Kenneth Roberts, ‘Neoliberalism and the Transformation of Populism in Latin America’, World Politics, vol. , no.  (), pp. –; Kurt Weyland, ‘Neopopulism and Neoliberalism in Latin America’, Studies in Comparative International Development, vol. , no.  (), pp. –; Weyland, ‘Neoliberal Populism in Latin America and Eastern Europe’, Comparative Politics, vol. , no.  (), pp. –. "# Pablo Gerchunoff and Juan Carlos Torre, ‘La polı!tica de liberalizacio! n econo! mica en la administracio! n de Menem’, Desarrollo EconoT mico, vol. , no.  (): –; Palermo and Novaro, PolıTtica y poder en el gobierno de Menem; Edward L. Gibson, ‘The Populist Road to Market Reform: Policy and Electoral Coalitions in Mexico and Argentina’, World Politics, vol. , no.  (April ), pp. –; Edward L. Gibson and Ernesto Calvo, ‘Electoral Coalitions and Market Reforms: Evidence from Argentina’, Paper presented at the XX International Congress of the Latin American Studies Association, Guadalajara, Mexico, – April, ; Sebastian Etchemendy and Vicente Palermo, ‘Conflicto y concertacio! n: Gobierno, Congreso y organizaciones de intere! s en la reforma laboral del primer gobierno de Menem’, Desarollo EconoT mico, vol. , no.  (), pp. –; Mariana Llanos, ‘El presidente, el congreso y la polı!tica de privatizaciones en la Argentina’, Desarrollo EconoT mico, vol. , No.  (). "$ Gibson, ‘The Populist Road’; Steven Levitsky, ‘From Laborism to Liberalism: Institutionalisation and Labor-Based Party Adaptation in Argentina, –’, Ph.D Dissertation (Department of Political Science, University of California, Berkeley, ); Javier Corrales, ‘Presidents, Ruling Parties, and Party Rules: A Theory on the Politics of Economic Reform in Latin America’, Comparative Politics, vol. , No.  (January ), pp. –. Local Party Structures in Argentine Peronism  evidence in this direction. It shows not only that the PJ organisation remained active in the s, but that the persistence of strong local organisations placed real limits on the Menem leadership’s capacity to transform the party from above. The article draws on data from two surveys carried out by the author in :() a survey of  local party branches (base units, or UBs), based on visits to the UBs and in-depth interviews with the activists who ran them; and () a survey of  party activists, based on a -point questionnaire distributed to activists in each of the surveyed UBs. The surveys were carried out in the Federal Capital, which is predominantly middle class, and the Greater Buenos Aires municipalities of La Matanza "% and Quilmes, which have larger working and lower class populations. UBs were selected so as to be as representative as possible of the socioeconomic, geographic, and internal factional make-up of each municipality. Although the Federal Capital, La Matanza, and Quilmes are not representative of the country as a whole, the organisational patterns observed in these districts were confirmed through interviews with party "& leaders and activists from several other cities in Greater Buenos Aires "' and other industrialised provinces, as well as through visits to a smaller number of base units in the provincial capital of San Miguel de Tucuman. Hence, there is reason to think that the findings of this study are "( generalisable, at least within urban Peronism.

An informal organisation: rethinking the Peronist party structure The anthropologists Gerlach and Hine once observed that scholars tend to assume that organisations are hierarchical and have a ‘well-defined ") chain of command’. Thus, ‘in the minds of many, the only alternative to a bureaucracy or a leader-centred organisation is no organisation at

"% La Matanza and Quilmes were selected because they are relatively representative of Greater Buenos Aires Peronism. The two districts lie near the median on several demographic and socioeconomic dimensions, and they each possess characteristics of the zone’s first and second ‘belts’. The first belt is a more established, industrialised zone with larger middle and working class populations, while the second belt is poorer, less developed, and populated by a larger number of internal migrants. Many of these migrants live in shantytowns or slums. Though strong in both belts, the PJ is particularly dominant in the second belt. "& These include Avellaneda, Beratazegui, Hurlingham, Ituzaingo, Jose! C. Paz, Lanus, "' and Tres de Febrero. These include Co! rdoba, Mendoza and Santa Fe. "( Interviews with leaders and activists from rural areas suggest that these patterns extend to those areas as well, although lack of data limits our ability to generalise with as much confidence about peripheral Peronism. ") Luther P. Gerlach and Virginia H. Hine, People, Power, Change: Movements of Social Transformation (New York, ), p. .  Steven Levitsky "* all’. Such has arguably been the case in studies of Peronism. Lacking the kind of disciplined, hierarchical, and bureaucratic organisation charac- teristic of many other working class parties, the Peronist party organisation has received almost no scholarly attention. The few existing analyses of the PJ organisation focus primarily on its formal, national- #! level leadership structure. This structure is indeed largely inoperative. A visit to the national party offices reveals a strikingly underdeveloped bureaucracy. The party’s top executive body, the National Council, lacks substantial resources or a professional staff, possesses little data on provincial party branches, and has virtually no record of its own activities #" prior to . Apart from a handful of custodial and low-level administrative staff, the national party headquarters are generally empty. Provincial and local party headquarters are even less active. In light of this bureaucratic weakness, scholars have often concluded that the PJ organisation is weak or even non-existent, and that party leaders maintain a personalistic and largely unmediated relationship with ## the Peronist rank-and-file. Thus, applying the well-known conceptual #$ frameworks of European scholars such as Kirchheimer and Panebianco, scholars have characterised the contemporary PJ as an ‘electoral- #% #& professional’ or ‘catch-all’ party. However, attention to the weakness of the PJ bureaucracy tends to obscure the powerful informal organisation that surrounds it. The Peronist organisation consists of a vast collection of informal networks that operate out of a range of different entities, including unions, cooperatives, clubs, soup kitchens, and often people’s homes. These informal entities

"* Gerlach and Hine, People, Power, Change,p.. #! Alberto Ciria, ‘Peronism and Political Structures, –’, in Ciria (ed), New Perspectives on Modern Argentina (Bloomington, ); Ciria, PolıTtica y cultura popular: la Argentina peronista, – (Buenos Aires, ); Carlota Jackisch, Los partidos polıTticos en AmeT rica Latina: Desarrollo, estructura y fundamentos programaT ticos. El caso argentino (Buenos Aires, ); Mustapic, ‘El Partido Justicialista’. #" When the newly-created Juan D. Pero! n Institute contacted party headquarters in  to invite National Council members to its public inauguration, party administrators were unable to even come up with the phone numbers or addresses of the body’s membership. ## Marcos Novaro, Pilotos de tormentas: crisis de representacioT n y personalizacioT n de la polıTtica en Argentina (‰ˆ‰–‰‰ƒ) (Buenos Aires, ), pp. –; Palermo and Novaro, PolıTtica y poder, pp. –; Weyland, ‘Neopopulism and Neoliberalism in Latin America’. #$ Otto Kirchheimer, ‘The Transformation of West European Party Systems’, in Joseph LaPalombara and Myron Weiner (eds), Political Parties and Political Development (Princeton, N.J., ), pp. –; Panebianco, Political Parties, pp. –. #% Novaro, ‘Peronismo y Menemismo’, pp. –. #& Eugenio Kvaternik, ‘El peronismo de los : un ana! lisis comparado’, Agora, no.  (), p. ; Mustapic, ‘El Partido Justicialista’, p. . Local Party Structures in Argentine Peronism  are self-organised and operated. They do not appear in party statutes, are rarely registered with local party authorities, and maintain near-total autonomy from the party bureaucracy. Yet they constitute the vast bulk of the PJ organisation. If, following Sartori, we define a political party as ‘any political group that presents at elections, and is capable of placing #' through elections, candidates for public office’, then all Peronist sub- units – formal or informal – that participate in electoral politics should be considered part of the party organisation. Studies of the PJ that focus on the party’s formal structure miss this informal infrastructure, and as a result, they miss the vast bulk of the party organisation. Rather then employ Panebianco’s distinction between ‘mass-bureau- #( cratic’ and ‘electoral-professional’ parties, then, it is perhaps more accurate to describe the PJ as an informal mass party. It is a mass party in that it maintains a powerful base-level infrastructure, extensive linkages to working and lower class society, and a large membership and activist base. It is informal in that Peronist sub-units organise themselves, lack a standard organisational structure, and are generally not integrated into (or subject to the discipline of) the party’s central bureaucracy.

The roots of informality: Peronism as a movement organisation The roots of the PJ’s informal structure lie in its distinctive history. #) Although Peronism originated as a charismatic party during the first Pero! n government (–), with a centralised, though non- bureaucratic, hierarchy based on the personalistic leadership of Juan #* Pero! n, the organisation changed considerably after Pero! n’s overthrow in . Banned and intermittently repressed throughout most of the – period, Peronism moved underground, surviving within trade unions, clandestine cadre organisations, and thousands of neighbourhood $! activist networks. However, unlike many other banned working class

#' Giovanni Sartori, Parties and Party Systems: A Framework for Analysis (Cambridge, #( ), p. . Panebianco, Political Parties, pp. –. #) On charismatic parties, see Panebianco (). #* For analyses of the early Peronist Party, see Walter Little ‘Party and State in Peronist Argentina’, Hispanic American Historical Review vol , no.  (), pp. –; Ciria, PolıTtica y cultura popular; Susana Elena Pont, Partido Laborista: Estado y sindicatos (Buenos Aires, ); Juan Carlos Torre, La vieja guardia sindical y PeroT n: Sobre los orıTgenes del peronismo (Buenos Aires, ); Mackinnon, ‘Sobre los orı!genes del partido peronista’. $! Activists formed ‘walking base units’, moving from house to house and holding meetings under the guise of barbecues or birthday parties. These groups engaged in a variety of clandestine activities, including study groups, ‘lightening meetings’ (in which activists would gather at a street corner, sing the Peronist March, and then flee), midnight graffiti brigades, masses for Evita, and literature distribution at soccer games.  Steven Levitsky parties (such as the French communists, German social democrats, and Venezuelan Democratic Action), which survived periods of repression by creating disciplined, hierarchical organisations, Peronism’s vertical links broke down after , and the organisation fell into a decentralised, semi-anarchic state. Initial acts of Peronist opposition were ‘spontaneous, atomised initiatives’, carried out in the ‘absence of a coherent, national $" leadership’. Peronists operated through self-constituted ‘commandos’ based on pre-existing union-based or neighbourhood friendship and $# family-based networks. Linkages among these local commandos were $$ ‘tenuous at best’, and the bodies created to coordinate their activities, such as Centre of Resistance Operations and the Peronist Grouping of $% Insurrectionary Resistance, were ineffective. Post- Peronism thus took on a segmented, decentralised structure that, following Gerlach and Hine, may be characterised as a ‘movement $& organisation’. Peronist sub-groups organised themselves autonomously from each other and from the central authorities. At the local level, activists formed neighbourhood ‘working groups’ or agrupaciones that maintained little systematic contact either with each other or with higher- level authorities. At the national level, Peronism was little more than a $' ‘loose federation of different groups loyal to Pero! n’, including unions, $( left and right wing paramilitary organisations, and numerous provincial $) ‘neo-Peronist’ parties. No single organisational order encompassed these sub-groups, and no central authority structure emerged with the capacity to coordinate their activities, discipline them, or even define who

$" Daniel James, Resistance and Integration: Peronism and the Argentine Working Class, ‰„†–‰‡† (New York, ), pp. –. $# James, Resistance and Integration,p.. $$ James, Resistance and Integration,p.. $% James, Resistance and Integration, pp. –. $& Gerlach and Hine, People, Power, Change. According to Gerlach and Hine, movement organisations may be distinguished from bureaucratic organisations in that they are segmented and decentralised (pp. –). They are segmented in that sub-units are largely autonomous from each other and do not interact regularly (pp. –). They are decentralised in that sub-units are not integrated into a central hierarchy. Hence, no central authority can ‘make decisions binding on all of the participants in the $' movement’ (p. ). James, Resistance and Integration,p.. $( These included Comando de OrganizacioT n and the Iron Guard on the right and the Montoneros, Descamisados and Peronist Armed Forces on the left. $) Neo-Peronist parties were Peronist organisations that competed in provincial elections under invented party labels, such as the Popular Union, the Populist Party, and Social Justice. Neo-peronist parties did not come together into a single national organisation, but rather were fragmented into autonomous (and competing) organisations. See Marı!a Fernanda Arias and Rau! l Garcı!a Heras, ‘Carisma disperso y rebelio! n: Los partidos neoperonistas’, in Samuel Amaral and Mariano Ben Plotkin (eds), PeroT n: del exilio al poder (Buenos Aires, ). Local Party Structures in Argentine Peronism  was or was not a Peronist. Although Pero! n remained the movement’s undisputed leader, his authority was limited to major decisions, and the bodies he created to represent him, such as the Superior and Coordinating Council and the Tactical Command, were routinely ignored by the unions, $* paramilitary groups, and provincial bosses. While Peronists who disobeyed Pero! n’s orders were sometimes ‘expelled’ from the movement, %! such expulsions were often ignored and were almost never permanent. After a brief return to power between  and , Peronism again fell into an anarchic state during the – dictatorship. The bulk of urban party activity migrated to the unions, but activists also worked within a range of clandestine organisations. Although base units were closed down, many continued to operate through informal ‘working %" groups’. Others ‘took refuge in non-governmental organisations’, such %# as sociedades de fomento, neighbourhood clubs, soup kitchens, and church %$ organisations. Still others worked out of front organisations such as %% ‘study centres’. Although little data exists on clandestine Peronist organisation during the Proceso, the number of activists who engaged in at least sporadic political activity appears to have been significant. Of the base units surveyed by the author in ,  per cent were led by an activist who worked in a Peronist organisation during the dictatorship. As a result of this clandestine work, when the military regime collapsed in , Peronism quickly re-emerged as a mass organisation. Base units mushroomed throughout the country – seemingly out of nowhere. By mid-, the PJ had signed up more than three million members, which %& was more than all other parties combined.

$* Arias and Garcı!a Heras, ‘Carisma Disperso y Rebellion;’ Miguel Bonasso, El presidente que no fue: Los archivos ocultos del peronismo (Buenos Aires, ). %! An example is Catamarca boss Vicente Saadi, who was expelled from the movement twice. In , Saadi was expelled after he disobeyed Pero! n’s order to back Arturo Frondizi in the  presidential election. Yet he retained control of Catamarca Peronism, and in , he was re-instated in the party. Similarly, when Pero! n ‘expelled’ the Montoneros in , Montonero leader Dardo Cabo declared, ‘No one has the right to throw us out. No one can fire us!’ (Quoted in Liliana De Riz, Retorno y derrumbe: El ultimo gobierno peronista (Buenos Aires, ), pp. –). %" Author’s interview with Anı!bal Stela, a local PJ leader in La Matanza,  July, . %# Non-governmental neighbourhood development centres. %$ In Tucuman, Peronists operated out of ‘neighbourhood centres’ in lower class areas, while in Federal Capital shantytowns, they created ‘neighbourhood juntas’ and soup kitchens. %% For example, a major Rı!o Negro faction worked through ‘Institutes of Rı!o Negro Studies’. %& ClarıTn,  April, ,p. and  July, ,p.. Given the party’s lack of access to state resources, this initial burst of party activism was based primarily on collective or non-material incentives. Although unions (which did have access to state resources) played a major role in financing party activity in the early s, a large number of  Steven Levitsky Unlike previous periods of civilian rule, during which major Peronist organisations ignored party activity, the PJ underwent an unprecedented process of ‘partyisation’ after . As elections came to be viewed as the only legitimate road to power, virtually all Peronist sub-units integrated themselves into party activity through participation in internal elections. %' Peronist unions invested heavily in party politics, as did former paramilitary organisations such as the Iron Guard, Comando de OrganizacioT n %( (C de O), the Peronist Youth (JP), and the Montoneros. By the mid-s, with the exception of the trade union realm, non-party Peronist activity had largely disappeared. The post- ‘partyisation’ process was not accompanied by a process of bureaucratisation, however. Rather than establishing a bureaucratic structure, the post- PJ retained key aspects of its movement organisation. Peronism re-emerged from the dictatorship in a bottom-up, semi-anarchic manner. Activists established their own base units without the approval (or even knowledge) of the party hierarchy. Not only did the party bureaucracy itself not create or finance base units, but it had no say over who could create them, how many were created, or where they were located. Moreover, while unions, ex-paramilitary organisations, and scores informal territorial networks all entered party activity in the s, they did not abandon their organisational forms or integrate themselves into the party bureaucracy. Rather, they remained self-organised, creating, financing, and operating their own base units. As a result, the PJ’s

activists were linked to the party by a shared identity and ideology. These activists were generally recruited through family and friendship networks and social organisations such as unions, cooperatives, clubs, and church groups. %' In a survey of  national unions carried out by the author,  reported participating in PJ politics in the s, and  reported placing members on PJ parliamentary lists or party leaderships. This evidence runs counter to James McGuire’s claim that two of the four largest Peronist union factions (the Ubaldinistas and the ‘’) did not participate in party activity in the s (McGuire, ‘Union Political Tactics and Democratic Consolidation in Alfonsı!n’s Argentina, –’, Latin American Research Review, vol. , no.  (), pp. –). While it is true that these factions did not work closely with the party at the national level, individual unions within these factions did in fact participate actively. %( Iron Guard networks established powerful territorial factions in urban districts such as the Federal Capital, Buenos Aires, and Santa Fe, winning six seats in congress and the governorship of Formosa. The Montoneros aligned with Catamarca boss Vicente Saadi to create a national party faction called Intransigence and Mobilisation, which was based on Montonero networks and financed largely by the Montoneros. C de O participated actively in internal elections in La Matanza and elected leader Alberto Brito Lima to congress. Even the most ‘movementist’ of Peronist Youth factions, such as Juan Carlos Dante Gullo’s Unified Peronist Youth, participated regularly in internal elections after . Local Party Structures in Argentine Peronism  national organisation remained a loose and heterogeneous collection of weak national factions, rump paramilitary organisations, labour organis- ations, and emerging provincial fiefdoms. Although the – Renovation period brought a degree of %) institutional order to the PJ, the reforms associated with this period were less far-reaching than is often believed. During this period, party reformers – called renovadores – took important steps to democratise the PJ internally (such as introducing direct elections for leadership and candidate selection) and paid unprecedented attention to its formal structure. Formal party organs such as the National Council met more frequently, the party began to keep records of its activities, and a greater effort was made to adhere to party statutes. Aside from the introduction of internal elections, however, the Renovation process did little to change the way the PJ actually functioned in practice. The renovadores failed to impose a standard organisational structure on the party and were unable to create an effective central bureaucracy capable of imposing discipline on lower level organisations. Consequently, Peronist sub-units remained informal and relatively autonomous.

The contemporary Peronist organisation: An informal mass party According to the party statutes that emerged from the  reform process, the contemporary PJ is formally structured as a European-style mass party, with a bureaucratic chain of command that runs from the National Council to provincial and municipal branches down to %* neighbourhood base units. In practice, however, the party more closely resembles what one Peronist mayor described as an ‘organised dis- &! organisation’. The PJ retains a powerful mass organisation with deep

%) Mustapic, ‘El Partido Justicialista’, pp. –; Marcos Novaro and Vicente Palermo, Los caminos de la centroizquierda: Dilemas y desafıTos del FREPASO y la Alianza (Buenos Aires, ), pp. –. %* The formal party hierarchy consists of four levels: () the national leadership, which includes the National Council and the party congress; () provincial-level branches, which are governed by provincial party councils; () municipal-level branches, which are governed by municipal councils; and () neighbourhood-level base units. The highest body of authority is the party congress, which has the power to determine the party programme, modify the party charter, and intervene provincial party leaderships. Day-to-day leadership of the party is carried out by a -member National Council and its -member Executive Board leaderships (Partido Justicialista, Carta OrgaT nica Nacional (Buenos Aires, ), articles –). &! Author’s interview with Juan Jose! Alvarez, mayor of Hurlingham,  July, . Local party leader Jose! Montenegro described the PJ as a ‘giant informal network’ (author’s interview,  Aug., ). According to Montenegro, ‘some Peronists work in the base units, others in their homes, and others in sociedades de fomento. But they are all part of  Steven Levitsky roots in working and lower class society, but these linkages continue to be non-bureaucratic, informal, and highly decentralised.

A mass organisation Although no modern day parties ‘encapsulate’ their members to the same &" degree that some turn-of-the-century European mass parties did, the PJ maintains what is by contemporary standards a powerful mass or- ganisation. This can be seen in several ways. First, it retains a large mass membership. Party membership stood at n million in , which &# represented a striking  per cent of the electorate. Indeed, the PJ’s membership-to-vote ratio of n exceeded those of postwar social &$ democratic parties in Austria, Germany and Sweden. Although the utility of such comparisons is limited by the fact that PJ membership entails a lower level of commitment than it does in most European mass &% parties, this extensive affiliation is nonetheless striking. Second, the PJ maintains a dense territorial infrastructure. Although the party’s failure to keep records of its base units makes it difficult to accurately measure the density of its organisation, evidence from La Matanza, Quilmes, and the provincial capital of San Miguel de Tucuman suggests that the PJ’s base-level infrastructure remains both extensive and densely organised. In , these three municipalities contained approxi- mately one UB per  residents and more than two UBs per square && kilometer. Third, the PJ remains deeply embedded in working and lower class society through linkages to a variety of (formal and informal)

the network. This network is always there, sometimes latent, but always ready to be activated’. &" Giovanni Sartori, ‘European Political Parties: The Case of Polarized Pluralism’, in Robert Dahl and D.E. Neubauer (eds), Readings in Modern Political Analysis (New York, ), pp. –; E. Spencer Wellhofer, ‘Strategies for Party Organisation and Voter Mobilisation: Britain, Norway, and Argentina’, Comparative Political Studies, vol. , no.  (), pp. –. &# Mark Jones, ‘Evaluating Argentina’s Presidential Democracy: –’, in Scott Mainwaring and Mathew Soberg Shugart (eds), Presidentialism and Democracy in Latin America (New York, ), p. . &$ Stefano Bartolini and Peter Mair, Identity, Competition, and Electoral Availability (Cambridge, ), p. . &% For example, Peronists do not pay regular dues. && In La Matanza, where there are approximately  UBs, there is an estimated one UB per  residents and . UBs per square kilometer. In Quilmes, where there are approximately  UBs, there is an estimated one UB per  residents and . UBs per square kilometer. (Based on demographic data in Informe de Coyuntura (La Plata, Nov.-Dec. ), p. ). In San Miguel, the PJ has approximately  UBs, which is equivalent to roughly one UB per  residents. Local Party Structures in Argentine Peronism  organisations. At the most basic level, local PJ organisations maintain extensive links to interpersonal networks in working and lower class neighbourhoods. In lower class zones, the ‘natural leaders’ or ‘problem &' solvers’ of the neighbourhood – are generally Peronists. Although many of these ‘natural leaders’ are not full time party activists, almost all of them maintain ties – through friends, neighbours, or relatives – to informal party networks. These ties are periodically activated both ‘from below’ and ‘from above’: neighbourhood ‘problem-solvers’ use them to gain access to government resources, while local punteros use them to &( recruit people for elections or other mobilisations. Local party organisations also maintain linkages to a range of social organisations. Historically, the most important of these have been unions. Although union influence in the PJ has declined considerably since the mid-s, the vast majority of unions remained active in local Peronist politics through the late s. Of  local unions surveyed by the author &) in ,  ( per cent) participated in party activity in . PJ organisations are also linked to a variety of urban social movements, such as squatters’ movements and shantytown organisations. In the Federal Capital, for example, the vast majority of shantytown organisations are run by PJ activists, and city-wide organisations of shantytown dwellers, such as the Shantytown Movement and the Social Front, maintained close ties to the PJ. In La Matanza, five of  UBs surveyed for this study were linked to squatter settlements, and the coordinator of the Roundtable of Squatter Settlements, which claimed to represent  shantytown &* organisations, is a member of the local PJ leadership. &' For an insightful account of Peronist ‘problem-solving networks’, see Javier Auyero, ‘The Politics of Survival: Problem-Solving Networks and Political Culture Among the Urban Poor in Contemporary Buenos Aires’, Ph.D dissertation, New School for Social Research, New York, . &( There is a dark underside to this social embeddedness. Because urban slum zones are frequently centres of illicit activity such as drug trafficking, prostitution, and gambling, Peronist networks are inevitably linked to these forms of organisation as well. Although verifiable data on illicit Peronist networks is difficult to obtain, it is widely believed that Peronist factions in La Matanza are linked to drug running, gambling, prostitution, and extortion networks. For example, networks of temporary workers in La Matanza’s Central Market complex, which are regularly mobilised by Peronist factions to paint graffiti and attend PJ mobilisations, are also rumored to be involved in drug trafficking and thug work, including the beating of a journalist who was writing a book on Buenos Aires governor Eduardo Duhalde. &) All of the surveyed unions were located in the Federal Capital, La Matanza, and Quilmes. &* Author’s interview with Rau! l Tuncho,  Aug., . Similarly, in the San Francisco de Solano zone of Quilmes, where , squatters carried out land invasions during the Proceso, the movement’s political organisation, Social Justice, joined the PJ in the early s (Luis Fara, ‘Luchas Reivindicativas urbanos en un contexto autoritario’, in Elisabeth Jelin (ed), Los nuevos movimientos sociales (Buenos Aires, ), pp. –).  Steven Levitsky Base-level PJ organisations are also linked to a variety of non- governmental organisations, including sociedades de fomento, school '! cooperatives, and soup kitchens. For example, shantytown organisers in the Federal Capital estimate that ‘seventy or eighty per cent’ of the city’s '"  soup kitchens are run by Peronists. Similar estimates have been made for the Greater Buenos Aires districts of Hurlingham, Lanus, and '# Quilmes. A smaller number of UBs maintain ties to church organis- ations. Finally, many Peronist organisations maintain ties to neigh- '$ bourhood and local clubs. Of particular importance are local soccer clubs (especially those in the second or ‘B’ division). Local PJ leaders often use soccer fan clubs to mobilise for rallies, paint graffiti, and, at times, intimidate opponents. Scores of party-club linkages exist in the Federal Capital and Greater Buenos Aires. For example, union leader Luis Barrionuevo’s control over the Chacarita soccer club helped him establish a powerful political base in Greater Buenos Aires. In La Matanza, the local party uses the Laferrere soccer fan club for mobilisations and graffiti '% painting. Party-club linkages are also common in the interior provinces. In Tucuman, for example, the two largest soccer clubs were controlled by '& Peronists in the late s. Overall, more than half (n per cent) of the UBs surveyed by the author showed evidence of a linkage to one or more social institutions, and more than a third (n per cent) were linked to two or more such '! In Quilmes, for example, most of the approximately  sociedades de fomento are run by Peronists. The Quilmes Federation of Sociedades de Fomento, which claims to represent  sociedades, linked itself to the PJ by creating the Justicialista Centre of Organised Communities and placing its president, Cornelio Melgares, in the local party leadership (author’s interviews with Cornelio Melgares,  April, , and Jose! Luis Saluzzi, Quilmes Director of Non-Governmental Entities,  Sept., ). '" Author’s interviews,  May, . '# The Hurlingham figure is based on the author’s interview with Hurlingham mayor Juan Jose! Alvarez,  July, . The Lanus figure is based on research by Javier Auyero (personal communication). The Quilmes figure is based on the author’s interview with Jose! Luis Saluzzi, Quilmes Director of Non-Governmental Entities,  Sept., . '$ In La Matanza, three of the largest clubs – Almirante Brown, Huracan, and Laferrere–have close ties to the PJ. In the Caballito neighbourhood in the capital, activists from the Peronist Unity Front have led the West Railroad Club and Italian Club since the s. '% In the Federal Capital, city council member Rau! l Padro! uses the Defenders of Belgrano fan club as an organisational base, while the UOM’s ties to the Nueva Chicago fan club strengthens its political presence in Mataderos and other southern neighbourhoods. In Lomas de Zamora, the PJ factions maintain close ties to the Banfield soccer club, and in Lanus, the local soccer club is closely linked to PJ mayor Manuel Quindimil. '& One of these clubs, Club Atletico, was run by PJ senator (and later governor) Julio Miranda. Similarly, in San Juan, control of the San Martin Club helped launch the political careers of PJ legislator Juan Jose! Chica Rodriguez and PJ governor Jorge Escobar. Local Party Structures in Argentine Peronism 

Table . Social linkages of base units surveyed in the federal capital and greater Buenos Aires* Number Percentage Base units with ties to at least one social organisation  n Base units with ties to at least two social organisations  n School cooperative  n Sociedad de fomento  n Neighbourhood club  n Self-help organisation  n Union  n Church organisation n Squatter organisation n Total  n * No information was available for eight of the surveyed base units. entities. This data is summarised in Table . Of the UBs surveyed, n per cent had ties to a school or child care cooperative, n per cent had ties to a local club, n per cent had ties to a sociedad de fomento, n per cent had ties to a soup kitchen or other self-help organisation, n per cent had ties to a union, n per cent had ties to a church organisation, and n per cent had ties to a squatter organisation.

An informal structure As they were throughout the – period, the PJ’s mass linkages remain informal and decentralised, rather than bureaucratic. The formal party structure is largely ignored, and power, resources, information, and even political careers pass through informal, self-organised sub-units with only weak and intermittent links both to each other and to the party bureaucracy. Neighbourhood-level organisation: base units. According to party statutes, '' base units constitute ‘the primary organ of the party’. They are the neighbourhood branches out of which activists operate. Formally, UBs must have at least  members and must be governed by a Base Unit '( Council that is elected every two years by the membership. In practice, however, they tend to be run by either a small group of activists or a single puntero (neighbourhood broker) and her or his inner circle of friends and family. Only n per cent of the UBs surveyed for this study held regular elections with leadership turnover. n per cent held nominal elections in which the same leader always won, and n per cent held no regular elections at all. '' Partido Justicialista, Carta OrgaT nica, article . '( Partido Justicialista, Carta OrgaT nica, articles , .  Steven Levitsky Formally, UBs are part of the party bureaucracy. According to party statutes, UBs register with, and fall under the direct authority of, local party councils. The local party council determines the jurisdiction of each UB, and in some districts it has the authority to intervene UBs that do not ') properly carry out their functions. In reality, however, UBs are autonomous from the party bureaucracy. They are not created by the party and their offices are not party property. Rather, they are created, privately, by the activists themselves. Anyone can open a UB at any time and in any place. Frequently, punteros establish UBs in their own homes and thus literally become their ‘owners’. As one local leader put it, We don’t have to ask, ‘may we open up a Peronist community centre?’ No. In Peronism you have the freedom to create what you want. No one tells you what '* to do. Of the surveyed UBs, n per cent were created ‘from below’ by activists or punteros, n per cent were created as outgrowths of existing UBs, n per cent were created ‘from above’ by local factions (or agrupaciones), and n per cent were created by unions. Not a single UB was created by the party bureaucracy. Not only do local party bureaucracies lack control over how many UBs exist or where they are located, but they often do not even have a record of the UBs under their jurisdiction. In Quilmes, party officials estimate (! that they have a record of about a third of the existing UBs. In La Matanza, party administrators claim to possess no records of existing UBs. Although party leaders have occasionally attempted to bring UBs under (" the control of the party bureaucracy, these efforts have repeatedly failed. For example, a  proposal in the Federal Capital to limit party offices to one officially recognised branch per ward was rejected as impossible to enforce. As one local leader put it, Who would finance and run these offices? And how are they going to close down the other base units if the base units are owned by the punteros? Are they going (# to throw the punteros out of their homes?’ UBs take a variety of organisational forms. While some take the form stipulated by the party statutes, others take the form of informal ‘working ') Partido Justicialista, Carta OrgaT nica, article ); Partido Justicialista de la Provincia de Buenos Aires, Carta OrgaT nica Provincial (La Plata, ), article . '* Author’s interview with Jose! Montenegro,  Aug., . (! Author’s interviews with PJ-Quilmes administrator Elba Quiroga ( Nov., ) and party president Jose! Rivela ( May, ). (" In Quilmes in the mid-s, for example, party president Roberto Morguen sought to permit only officially recognised UBs to operate in each neighbourhood, but because the party had no means of enforcing the policy, the policy was widely ignored (author’s interviews with Roberto Morguen ( May, ) and Jose! Luis Saluzzi ( Sept., )). (# Author’s interview with PJ-Federal Capital treasurer Rau! l Roa,  Nov., . Local Party Structures in Argentine Peronism  groups’, which operate out of activists’ homes, without a sign outside or any kind of formal recognition from the party. Legitimated by Pero! n’s proscription era motto that ‘every house is a base unit’, working groups are formed by activists without the numbers or resources to maintain an office, or by punteros who prefer informality because it gives them more ‘room for manoeuvre’. Other UBs operate out of civic associations such ($ as neighbourhood cooperatives, community centres, or soup kitchens. Most of these entities function just like UBs, signing up party members, (% competing in internal elections, and campaigning in general elections. According to a local leader who runs the ‘Companions’ UB in the capital, We put on different hats. One day we are a base unit; the next day a child care centre, and the next day the civic association. But we always have the same (& Peronist ideology. In some areas, a majority of UBs operate out of civic associations. In San Miguel (province of Tucuman) for example, most UBs function as ‘neighbourhood centres’, and in the province of Santa Cruz, a large number of UBs are organised into ‘community centres’. Finally, some UBs are actually neighbourhood annexes of local party factions. Neighbourhood annexes differ from other UBs in that they are created (and often staffed) from the outside, rather than by neighbourhood activists. Like working groups and civic associations, however, they are not sanctioned by, or subject to the authority of, local party authorities. PJ organisations thus consist of a heterogeneous mix of UBs, working groups, civic organisations, and non-profits. For example, in the Lugano neighbourhood in the Federal Capital, the MOVIP faction includes a child care centre, a comedor, a ‘mother’s centre’, a community centre, and several working groups and UBs, while the Loyalty faction in Quilmes includes a church group, a mother’s association, a children’s rights group, and several community centres. Table  shows the distribution of organisational forms taken by the UBs surveyed for this study. Less than half (n per cent) took the form stipulated by the party charter; n per cent were informal working groups; n per cent operated out of civic associations; and n per cent were neighbourhood annexes sponsored ($ Such arrangements are generally an effort to gain access to state subsidies or to administer state-financed social programmes. They also represent an effort to appeal to a broader, non-Peronist constituency. (% As one activist put it, ‘We work for nine months out of the year as a civic association, and then at election time we turn ourselves into a base unit’ (author’s interview,  March, ). (& Author’s interview with Liliana Monteverde,  June, . Similarly, an activist whose UB was transformed into a comedor said, ‘we had to take down the Peronist sign to get funding, but everyone knows it is still a base unit’ (author’s interview,  March, ).  Steven Levitsky

Table . The organisational form of surveyed base units Organisational Form Number Percentage Formal Base Unit  n Informal ‘Working Group’  n Civic Association\Non-Profit  n Annex of AgrupacioT n  n Union n Total  n by local factions, or agrupaciones. A small number (n per cent) operated out of unions. Municipal-level organisation: agrupaciones. UBs are linked to the party through informal organisations called agrupaciones. Agrupaciones are clusters of UBs that compete for power in the party at the local level. They may be led by government officials, city council members, provincial or national legislators, unions, or outside political entrepreneurs seeking to build a base. They range in size from four or five UBs to more than , although most contain between  and . Two or three dozen agrupaciones may exist in each municipality, but usually only a handful are politically influential. In their competition for votes, agrupaciones build alliances with punteros throughout the municipality. In exchange for their support, punteros seek direct financing for their UBs, material resources to distribute to their constituents, and, whenever possible, government jobs. Because those in the best position to offer such resources are public office holders, most agrupaciones are held together by state patronage. Public officials use money from kickbacks, resources from the government agencies they run, and low level government jobs to finance their agrupaciones. Food and medicine from social welfare ministries are routinely diverted to UBs, and punteros are widely employed in government (' offices. Of the UBs surveyed for this study,  per cent belonged to an agrupacioT n, and most of the others were in transition from one agrupacioT n to another. Agrupaciones are thoroughly informal organisations. They are not mentioned in the party charter, and local party offices generally keep no record of them. Because they organise and finance themselves, agrupaciones enjoy substantial autonomy from the party bureaucracy. They do not have to conform to either the dictates of party statutes or the decisions of local party authorities. This organisational autonomy is legitimated by Peronism’s ‘movementist’ tradition. Because the Peronist ‘movement’ is understood to be broader than the party, and because the movement lacks (' Punteros commonly receive fictitious government contracts that enable them to earn full time salaries while they work in the UBs. Local Party Structures in Argentine Peronism  any kind of (formal or informal) structure or hierarchy, Peronists often conceive of their agrupaciones as lying outside the party, but inside the (( movement. Thus, the PJ’s boundaries are fluid: agrupaciones float in and () out of the party with relative ease. It is the agrupaciones, not the party bureaucracy, that link the party to its mass base. The vast bulk of party resources – particularly patronage and (* other unregulated forms of finance – pass through the agrupaciones, and it is the agrupaciones, not the party bureaucracy, that finance and maintain )! the UBs. Of the UBs surveyed by the author, n per cent received most or all of their resources from an agrupacioT n, while n per cent were either self-financed or were financed through private sector donations. Not a single UB was financed by the party bureaucracy. Agrupaciones also carry out the bulk of the party’s mobilisational work. Punteros mobilise their supporters for activities sponsored by the agrupacioT n, carry out instructions that are channeled through the agrupacioT n, and distribute material goods and literature supplied by the agrupacioT n. Punteros get the bulk of their information from assemblies run by their agrupacioT n, and to the extent that UBs are able to channel demands upward through the party, they do so via the agrupacioT n. By contrast, UBs rarely (( Thus, despite the fact that it participates in internal elections and campaigns for the PJ in general elections, the ex-paramiliary group Comando de OrganizacioT n in La Matanza considers itself ‘part of the movement, but not part of the party’ and therefore in no way subject to the discipline of the local party (author’s interview with Alberto Brito Lima,  April, ). Similarly, a Quilmes city councilman Reymundo Gonzales claimed he worked in an agrupacioT n so as to ‘avoid impositions by the party’ (author’s interview,  June ). () In the Federal Capital, for example, Peronist Victory (VP) left the party in  to back the mayoral candidacy of Gustavo Beliz of the New Leadership party. Over the next two years, VP maintained ‘one foot in and one foot out of the party’, calling itself a ‘Peronist agrupacioT n outside the structure of the PJ’ (author’s interviews with VP leaders Jorge Arguello ( May, ) and Victor Pandolfi ( June, )). In ,VP aligned with Domingo Cavallo’s Action for the Republic party. In , VP returned to the PJ to join to back Eduardo Duhalde’s presidential campaign. (* Although public finance is channeled through the party bureaucracies, this amounts to a relatively small percentage of party finance. Most private donations – and, crucially, all patronage resources – are channeled through the agrupaciones. )! According to former Federal Capital party treasurer Rau! l Roa, ‘The party bureaucracy just maintains the headquarters, which is nothing more than an office and the employees that clean the place. The rest of the organisation is financed and coordinated by the leaders of the agrupaciones. The party’s real infrastructure is in the hands of the agrupaciones’ (author’s interview,  May, ). Indeed, whereas the headquarters of the PJ Metropolitan Council in the Federal Capital is generally staffed only by cleaning personnel, agrupaciones such as the Peronist Unity Front and Liberators of America maintain computer-equipped headquarters with full time staff. Similarly, in La Matanza, the PJ’s de facto headquarters in the s was the office of the dominant Militancy and Renovation (MyRP) agrupacioT n. Known as La Casona, the MyRP headquarters was better financed, better staffed, and more frequently visited than the party headquarters.  Steven Levitsky participate in events sponsored by the local party leadership. Even political careers are channeled through agrupaciones. Because leadership recruitment and candidate selection is done almost entirely by the agrupaciones, emerging politicians invest in agrupaciones rather than build careers in the party bureaucracy. Agrupaciones also have a greater capacity to impose discipline on local leaders than does the party bureaucracy. For example, despite the fact that the Federal Capital party statutes stipulate that elected officials must )" contribute  per cent of their salaries to the local party, only three of )# six congressional deputies did so in . By contrast, every one of the elected officials belonging to Liberators of America and FUP regularly )$ contributed  per cent of their salaries to their agrupaciones. A similar situation exists in Quilmes, where local party president Jose! Rivela claims that ‘less than half’ of elected officials meet their obligation to contribute )% five per cent of their salaries to the party. According to Rivela, local Peronists ignore the party leadership … . They respond to the agrupaciones because that is where the money comes from … . The agrupaciones can provide them with something – money to pay the rent, or food or blankets to give away. The party )& can’t offer anything, so no one pays any attention to it. Provincial and national party organisation. At the provincial level, agrupaciones aggregate into competing factions or lıTneas internas, which are almost always led by public officials, such as governors, national or )' provincial cabinet members, or legislators. The primary currency of exchange between lıTneas internas and agrupaciones is patronage: agrupaciones exchange votes for positions in provincial governments or on national or provincial legislative lists. LıTneas internas vary in their organisational structures. Some are loosely organised factions that emerge for internal elections and then disappear. Others, such as the Peronist Renovation Movement in Santa Cruz, the Orange List in Mendoza, and Peronist Convergence in La Pampa, endured for more than a decade. Although a )" Partido Justicialista Metropolitana, Carta OrgaT nica Metropolitana, (Buenos Aires, ), article . )# Author’s interview with Federal Capital party treasurer Rau! l Roa,  May, . )$ Author’s interviews with Rau! l Roa of FUP and Victor Columbano of Liberators of )% America,  May, . Author’s interview,  May, . )& Author’s interview,  May, . )' In Buenos Aires, the party was dominated by two factions in the s: the Federal League, which was run by Chamber of Deputies president Alberto Pierri and provincial public works minister Hugo Toledo, and the Buenos Aires Peronist League, led by provincial chamber of deputies president Osvaldo Mercuri. Both factions were loyal to governor Eduardo Duhalde. In Tucuman, the two dominant lıTneas internas in the s – ‘True Peronism’ and ‘Peronism of Hope’ – were led by the party’s two national senators, Olijela Rivas and Julio Miranda. Local Party Structures in Argentine Peronism  few lıTneas internas are well-institutionalised and maintain relatively coherent )( programmatic profiles, most are highly personalistic, often taking on the )) name of their leader. Power, resources, and careers pass through the lıTneas internas, rather than formal party hierarchies, and as a result, provincial party bureaucracies tend to be weak. For example, the Buenos Aires party branch, which has a membership of more than one million, was only open for three half days every week in  and had no full-time staff. The provincial council rarely met, and when it did, it often had to call acting )* party president Alberto Pierri to have the office unlocked. By contrast, the party’s two dominant factions, the Federal League and the Buenos Aires Peronist League kept detailed records of local-level party activity, organised and financed the party’s campaigns, mobilised Peronists for rallies, and disciplined local organisations. In Tucuman, when the factions stopped funding the party in , the provincial party headquarters was left for several weeks without phone service, running water, or funds to *! finance its congressional campaign. Hence, without control over the lıTneas internas, control of the bureaucracy means very little. Indeed, when the party’s formal leadership is not aligned with the dominant factions, it *" becomes virtually powerless. Through the s, the PJ lacked an effective structure – even an informal one – at the national level. As is the case at the local and provincial levels, the national party bureaucracy is largely inoperative. To date, however, no national-level equivalents to agrupaciones and lıTneas *# internas have emerged. Provincial bosses are not linked together horizontally or vertically integrated into a central hierarchy, but rather tend to remain entrenched in their own fiefdoms. )( Examples include Peronist Renovation Movement in Santa Cruz, Eva Pero! nin Formosa, and, more ambiguously, the Orange List in Mendoza. )) Examples include Jaurismo (after governor Carlos Juarez) in Santiago del Estero, Romerismo (after governor Juan Carlos Romero) in Salta, and Saadismo (after Vicente and Ramon Saadi) in Catamarca. )* Author’s interview with Oscar Guida, a member of the PJ-Buenos Aires provincial council,  Nov., . *! Author’s interview with PJ-Tucuman president Amado Juri,  Dec., . *" This was clearly seen in the case of the  referendum in Buenos Aires on a set of reforms to the provincial constitution that would have permitted governor Antonio Cafiero to run for re-election. When the measure was handily defeated, with many Peronists voting against it, the outcome was taken as a major surprise given that Cafiero controlled the party ‘apparatus’. Yet two of the party’s three major lıTneas internas, the Federal League and ‘Menem Leadership’, did not work in favour of the measure and, in many places, quietly worked against it. By August , when the referendum was held, the Federal League had grown substantially in the province, leaving Cafiero with a minority of the ‘real’ party. *# Only one national faction – the renovadores – emerged in the post- period, and it disintegrated soon after its defeat in the  internal elections.  Steven Levitsky Due to this informal and segmented structure, the PJ is substantially more decentralised than is often believed. Lacking an effective central bureaucracy, the PJ functions like an organisational ‘big tent’, containing within it diverse and often contradictory elements. Far from the centralised *$ or ‘verticalist’ tradition with which Peronism is often associated, the relationship between higher and lower level PJ authorities is actually *% closer to one of mutual autonomy. On the one hand, in the absence of effective horizontal links, Peronist sub-units have difficulty acting collectively to check the power of central leaders, which gives PJ leaders *& substantial autonomy from lower level authorities. On the other hand, party leaders lack effective mechanisms with which to impose discipline on party sub-units. Consequently, unlike centralised mass parties such as the Chilean communist party or the Venezuelan AD, in which sub-units must strictly adhere to the national party line or face expulsion, the PJ exhibits a substantial degree of internal tolerance and diversity. As one PJ legislator put it, In other parties, everything one does has to be approved by the party hierarchy … . Your discourse has to conform to certain party standards. In *' Peronism, none of that is true. You can do or say whatever you want.

The PJ under Menem: The surprising resilience of local and provincial party organisations By virtually any measure, PJ underwent a set of radical changes in the s. Under the leadership of Carlos Menem, a Peronist government dismantled the statist economic model that had been in place since the s and established one of the region’s most open economies. Early analyses depicted this reform process as a kind of ‘revolution from above’, imposed by a powerful president acting at the margins of the PJ, *( major interest groups, and the legislature. Argentina was thus widely viewed as a case of ‘neo-populism’, in which the president circumvents *) parties in favour of direct, unmediated appeals. In line with these analyses, the PJ was viewed as having been eviscerated or transformed from above in the s. According to Marcos Novaro, the PJ ‘was *$ See Ciria, PolıTtica y cultura popular. *% Eldersveld used the terms ‘stratarchy’ and ‘reciprocal deference’ to describe a similar phenomenon in U.S. parties. See Samuel D. Eldersveld, Political Parties: A Behavioral Analysis (New York, ), pp. –. *& Levitsky, ‘From Laborism to Liberalism’, pp. –. *' Author’s interview with congressional deputy Fernando Maurette,  July, . *( O’Donnell, ‘Delegative Democracy’. *) Roberts, ‘Neoliberalism and the Transformation of Populism;’ Weyland, ‘Neo- populism and Neoliberalism in Latin America;’ Weyland, ‘Neoliberal Populism in Latin America and Eastern Europe’. Local Party Structures in Argentine Peronism  completely reorganised, from the highest positions to each one of its local ** chapters’. Menem, he argues, ‘reduced the role of the party organisation to a minimum’, such that the PJ came to function largely as an ‘electoral "!! committee’. From this perspective, the national party leadership "!" ‘functioned as a … mechanism of control of provincial leaderships’, allowing Menem to impose strategies on provincial branches and replace "!# local party leaders and candidates with media-friendly ‘outsiders’. This section offers a somewhat different picture of the PJ-Menem relationship. It argues that unlike ‘neo-populist’ leaders such as Collor and Fujimori, President Menem’s relationship to the Peronist rank-and- file was always mediated by strong, semi-autonomous local organisations. These base-level organisations provided the governing party with a range of political benefits in the s. Yet they also proved double-edged, for they limited Menem’s capacity to impose leaders, candidates, and strategies on lower level branches. As a result, local and provincial Peronist organisations remained surprisingly ‘un-Menemised’ at the end of the s. Base-level party activity in the ‰‰€s Recent studies have emphasised the central role of the Peronist party in building and maintaining support for the Menem government’s reform "!$ programme. Gibson and Calvo, for example, have noted the importance of ‘well-established networks of political support’ in delivering Peronist "!% votes in the provinces. Although these authors focus on the peripheral provinces, it is clear that local organisations were critical to maintaining mass support in poor urban areas – such as Greater Buenos Aires and Greater Rosario – as well. Not only did the PJ’s vast infrastructure of UBs, unions, soup kitchens, clubs, and informal social networks yield vast human and organisational resources for campaigns, but it also provided channels for patronage distribution, policy implementation, social and cultural contact, and (albeit with less frequency) political participation and ** Marcos Novaro, ‘Shifting Alliances: Party Politics in Argentina’, NACLA Report on the Americas, vol. , no.  (), p. . "!! Novaro, ‘Menemism and Peronism’, pp. –. Similarly, McGuire cites an Argentine newspaper column that claims that ‘the committees of the defeated UCR show more life today than the base units of the victorious Partido Justicialista’ (James McGuire, Peronism without PeroT n: Unions, Parties, and Democracy in Argentina (Stanford, ), "!" p. ). Palermo and Novaro, PolıTtica y Poder,p.. "!# Novaro, ‘Menemismo y Peronismo, pp. –; Maria de los Angeles Yannuzzi, La modernizacioT n conservadora: El peronismo de los ‰€ (Buenos Aires, ), pp. –; McGuire, Peronism without PeroT n, pp. –) "!$ Gibson, ‘The Populist Road to Market Reform;’ Gibson and Calvo, ‘Electoral Coalitions and Market Reforms;’ Corrales, ‘Presidents, Ruling Parties, and Party Rules’. "!% Gibson and Calvo, ‘Electoral Coalitions and Market Reforms’.  Steven Levitsky demand-making. For example, during the – hyperinflationary crisis, tens of thousands of party activists worked to dampen working and lower class protest in response to the hyperinflationary crisis and the government’s austerity measures. This was done through persuasion, the "!& physical expulsion of leftist activists from their neighbourhoods, and a variety of neighbourhood-based emergency social welfare measures. In La Matanza, for example, Peronist activists reportedly operated more than "!'  city-sponsored soup kitchens in . Peronists also operated "!( scores of soup kitchens in the shantytowns of the Federal Capital. In the city’s first ward, activists from the ‘United or Dominated’ UB claim to have distributed food to  families, going door to door in their neighbourhood to find out who was in need. In the nd ward, punteros organised collective soup kitchens in which families with food provided for those who lacked it. Local Peronist organisations engaged in a variety of other political, social, and cultural activities in the s. For example, party activists played a major role in the delivery of social assistance in lower income neighbourhoods. Of the base units surveyed for this study,  per cent engaged in some form of social assistance. Although scholarly and journalistic accounts of this material goods distribution tend to portray it "!) as naked clientelism, virtually no research has been done on what UBs "!* actually do. Indeed, as recent work by Javier Auyero demonstrates, the picture is somewhat more complex. Much Peronist base-level social assistance is in fact clientelistic. As Table  shows, more than two-thirds (n per cent) of the UBs surveyed for this study engaged in the direct distribution of food or medicine, and nearly a quarter (n per cent) of the UBs regularly provided jobs for their constituents. Yet UBs also provide a range of other social welfare services –

"!& Author’s interview with La Matanza city council member Abraham Delgado,  Nov., , and activists Cacho Ines, and Graciela Diaz,  June, . "!' Author’s interview with Anı!bal Stela, who served as vice-president of the PJ in La Matanza during the hyperinflationary period,  June, . "!( Based on the author’s interviews with PJ activists Reinaldo Mendoza ( Oct., ), Eugenio Lammardo ( June, ), Mate Ocampo ( March, ), Carlos Racedo ( March, ), and Ana Suppa ( July, ). "!) See Judy Lawton, ‘Clientelist Politics and Peronism in the Squatter Settlements of Greater Buenos Aires: Squatters’ Views on Politics and Society’, paper delivered at the XVIII International Congress of the Latin American Studies Association, Atlanta, March ; Hernan Lo! pez Echague, El otro: Una biografıTa de Eduardo Duhalde (Buenos Aires, ); Daniel Otero, El entorno: La trama ıTntima del aparato duhaldista y sus punteros (Buenos Aires, ); PaT gina\‚,  Oct., ,p.. "!* Auyero, ‘The Politics of Survival’ and ‘Evita como performance: Mediacio! ny resolucio! n de problemas entre los pobres urbanos de Gran Buenos Aires’, in Javier Auyero (ed), ¿Favores por votos? Estudios sobre clientelismo polıTtico contemporaT neo (Buenos Aires, ). Local Party Structures in Argentine Peronism 

Table . Social welfare activities of surveyed base units, by social class of neighbourhood* (percentages in brackets) Lower class Working class Middle class Total Activity (n l ) (n l ) (n l ) (n l ) General social assistance     (n) (n) (n) (n) Direct distribution of     food or medicine (n) (n) (n) (n) Provision of government     jobs (n) (n) (n) (n) Programmes for children     (n) (n) (n) (n) Programmes for the     elderly (n) (n) (n) (n) Legal assistance     (n) (n) (n) (n) Neighbourhood social     and cultural events (n) (n) (n) (n) Regular delivery of     particularistic favours (n) (n) (n) (n) Neighbourhood     improvement (n) (n) (n) (n) * Categorisations of social class are author’s judgements, based on firsthand observation and interviews with neighbourhood activists. including medical and legal services, child care, job training, and programmes for the elderly – that are less directly tied to a political exchange. For example, the ‘Reconquest’ UB in the Federal Capital offers school tutoring for children, computer classes and a job location programme for adults, and a retirees centre for the elderly, the ‘October ’ UB in Quilmes runs a health clinic and provides school uniforms for a local school, and the ‘Juan Manuel de Rosas’ UB in La Matanza founded a child care centre, organised a neighbourhood youth soccer team, and runs a retirees centre. Of the UBs surveyed for this study, n per cent provided regular activities for children, n per cent offered programmes for the elderly, and n per cent provided free legal assistance for low- income residents. UBs also engage in a variety of social and cultural activities. For example, the ‘Nelson Calvi Peronist House’ in the capital runs a youth soccer programme, holds monthly parties to celebrate neighbours’ birthdays, and throws a well-attended annual Children’s Day party; the ‘Ramon Carillo’ UB in Quilmes organises barbecues for workers from a nearby factory; and the ‘Menem Leadership’ UB in La Matanza offers martial arts, movie nights, and dance programmes for teenagers. Overall,  Steven Levitsky n per cent of the UBs surveyed regularly organised social and cultural activities in their neighbourhoods. A smaller number of UBs organise specifically Peronist cultural activities, such as masses for Evita and the celebration of Peronist holidays ""! such as Evita’s birthday, October , and the ‘Day of the Activist’. About a third of the surveyed UBs (n per cent) surveyed regularly """ sponsored such Peronist cultural activities. A few UBs (n per cent) continued to teach Peronist ‘doctrine’, either through classes, reading groups, or the distribution of Pero! n’s writings. Peronist base organisations play a critical role in linking working and lower class citisens to the state. Many UBs participate directly in the implementation of government social programmes. Although such politicisation is often viewed as a corrupt and inefficient distortion of state ""# policy, in many lower class areas, the state bureaucracy is so weak that party networks are a more effective means of reaching the population. An example is the Pierri Law, a programme through which tens of thousands of families received legal titles for their properties. Because many local residents were unaware of the requirements of the programme, and because local governments lacked the resources to carry out an extensive grassroots campaign, Peronist activists frequently provided the legwork, ""$ going door-to-door and helping residents do the paperwork. Another example is the Life Plan, which distributes a daily ration of eggs, milk and other basic goods to nearly , people through a network of , volunteer manzaneras, or block workers. Although the programme is officially non-partisan (manzaneras are chosen though community organis- ations), the vast majority of manzaneras are Peronist, and most are linked ""% to the party through informal Peronist networks. As Table  shows,

""! The Day of the Activist, Nov. , marks the day in which Pero! n returned from exile in . """ For example, the ‘Companions’ UB in the Federal Capital carried out an ‘Evita campaign’ in  in order to ‘reactivate the memory’, while the ‘Juan Manuel de Rosas’ UB in La Matanza organised an Evita Day, in which dozens of women were invited to discuss ‘what Evita means to me’. ""# See, for example, the various publications of the Instituto Bonaerense de Ana! lisis y Proyectos (IBAP), as well as Lo! pez Echague, El otro, pp. –. ""$ Another example is the UGE plan, a Buenos Aires programme in which unemployed residents are paid to pave the streets of their neighbourhoods. To qualify for the programme, neighbourhoods must form cooperatives, collect a quota of signatures, and submit an application. These tasks are often performed by UBs, and as a result, punteros frequently run the UGE programme in their neighbourhoods. ""% For example, in his research in a Lanus shantytown, Auyero found that  of the zone’s  Manzaneras were recruited from Peronist networks (Auyero, ‘The Politics of Survival’, p. ). In La Matanza, local Life Plan director Mario Ferreri acknowledged that a ‘majority’ of the non-governmental organisations from which Manzaneras are selected are ‘run by Peronists’ (author’s interview,  September, ). Local Party Structures in Argentine Peronism 

Table . Base unit participation in government social programmes in greater Buenos Aires* Number Government Programme (n l ) Percentage Participate in at least one  n programme Life Plan n Public works jobs  n programmes Pierri Law  n UGE (street pavement)  n Soup kitchen n * Because these programmes are sponsored by the provincial government of Buenos Aires, base units from the Federal Capital, which lies outside the province of Buenos Aires, were excluded from the sample. Three additional base units were excluded due to insufficient information.

n per cent of the Greater Buenos Aires UBs surveyed for this study participated in at least one government programme. PJ activists also play an important role ‘from below’ in providing residents of lower class neighbourhoods with access to the state. Where the PJ controls the local government, activists use their ties to public officials to act as a ‘nexus between the neighbourhood and the city ""& government’. Thus, local PJ organisations serve as ‘problem solving ""' networks’, obtaining wheelchairs, disability pensions, scholarships, funeral expenses, and odd jobs for working and lower class residents who lack alternative sources of social assistance. Not all problem-solving networks are particularistic. Activists also employ them to obtain collective goods and services for their neighbour- hoods. In La Matanza, for example, the ‘Pero! n and Evita’ UB played a major role in bringing street lights, paved roads, and bus service to a neighbouring squatter settlement. In the capital, the ‘October ’UB brought a child care centre, a gymnasium, and a computer-equipped job training centre to the Ciudad Oculta shantytown, the Nelson Calvi Peronist House helped to install a sewage system in the Soldati neighbourhood, and the ‘United or Dominated’ UB began teenage pregnancy and battered women’s programmes in the city’s first ward. Overall, n per cent of the UBs surveyed for this study – and  per cent in lower class neighbour- hoods – engaged in such neighbourhood improvement work (see Table ). Peronist activists also engage in political fights in defense of their neighbourhoods. In La Matanza, for example, the ‘Pero! n and Evita’ UB ""& Author’s interview with La Matanza activist Tina Blanco,  May, . ""' Taken from Auyero, ‘The Politics of Survival’.  Steven Levitsky represented the neighbourhood when the regional electric company sought to collect back debts that residents could not pay. In the Federal Capital, the ‘United or Dominated’ UB has defended a nearby squatter settlement against city government efforts to remove it. In Quilmes, UBs from the Loyalty faction organised protests that forced the regional water company to restore service to the neighbourhood after it had been cut off for non-payment of debts, successfully pressured the city government to increase police patrols in the neighbourhood, and led a petition drive to pressure the provincial government to take action to prevent flooding from a nearby river. Finally, a minority of UBs function as channels of grassroots participation by creating arenas for debate or holding regular meetings with local politicians. In Quilmes, for example, the ‘Cooperativism and Social Justice’ UB organised a day-long workshop in which women from poor neighbourhoods discussed their socio-economic problems with local leaders. In the Federal Capital, the ‘Hour of the People’ UB holds monthly public lectures and debates on issues such as the future of the government’s economic programme, social policy, and labour market reform. Also in the capital, the Justicialista Victory UB holds well- attended weekly political meetings with city councilwomen Kelly Olmos. Overall, n per cent of the surveyed UBs showed evidence of some kind of non-electoral political activity, and n per cent showed evidence of high and sustained levels of political activity. Although the political impact of this base-level party activity is difficult to measure, it undoubtedly helped to reinforce and sustain the Peronist subculture and identity in the s. Though weakened by generational ""( change and the penetration of mass media technologies, a common body of language, symbols, traditions, practices, and beliefs continues to unite Peronists of different ages, regions, social backgrounds, and ideologies. For many working and lower class voters, the Peronist identity extends beyond party politics into the social and cultural "") realms. For these voters, Peronism continues to be less of a party choice ""* than an encompassing identity. The persistence of this identity raises the threshold at which such voters decide to abandon the PJ. Indeed, as Pierre Ostiguy has shown, the PJ’s traditional electorate remained relatively stable in the s, despite the fact that many traditional ""( Oscar Landi, ‘Outsiders, Nuevos Caudillos y Medias Polı!ticas’, in Carina Perelli, Sonia Picado, and Daniel Zoviatto (eds), Partidos y clase polıTtica en AmeT rica Latina en los ‰€ (San Jose! , ); Silvio Waisbord, El gran desfile: Campang as electorales y medios de comunicacioT n en Argentina (Buenos Aires, ). "") Pierre Ostiguy, ‘Peronism and Anti-Peronism: Class-Cultural Cleavages and Political Identity in Argentina’, Ph.D dissertation, University of California at Berkeley, . ""* Ostiguy, ‘Peronism and Anti-Peronism’, pp. –. Local Party Structures in Argentine Peronism 

Table . Activists’ views on the menem economic programme* View of the government’s economic policies Percentage Agree with the neoliberal programme n Disagree with the neoliberal programme but back the n government out of loyalty Disagree with the neoliberal programme but believe that no n other option existed Disagree with major aspects of the neoliberal programme and n believe other options were available Fully oppose the neoliberal programme as a betrayal of Peronist n ideals Total n

Activist Views of Government Policies Toward Business, Unions, and Workers Business Unions Workers Too Favourable nnn Correct nnn Too Unfavourable nnn Total n n n

The next Peronist government should … Percentage Maintain the Menem economic model n Maintain the Menem model, but with more social justice n Return to the roots of Peronism n * Based on author’s survey of  PJ activists in the Federal Capital, La Matanza, and Quilmes in .

Peronists did not agree with the Menem government’s economic "#! policies. Local autonomy and the stability of the PJ activist base The stability of the PJ activist base in the s presents somewhat of a puzzle. Party activists, or at least an important subset of activists, are "#" generally though to be more ideologically-driven than party leaders. If that is the case, then we should expect that PJ activists, who have historically been almost uniformly anti-liberal, would have abandoned the party in droves in response to the Menem government’s neoliberal turn. Data from the  activist survey suggests that PJ activists were in fact quite critical of the Menem programme. As Table  shows, more than

"#! Ostiguy, ‘Peronism and Anti-Peronism’. "#" John D. May, ‘Opinion Structure of Political Parties: The Special Law of Curvilinear Disparity’, Political Studies vol. , no.  (), pp. –; Panebianco, Political Parties, pp. –; Kaare Strom, ‘A Behavioral Theory of Competitive Political Parties’, American Journal of Political Science, vol. , no.  (), pp. –; Alan Ware, ‘Activist-Leader Relations and the Structure of Political Parties’, British Journal of Political Science vol.  (January ), pp. –.  Steven Levitsky two-thirds of surveyed activists opposed either part or all of the Menem programme. Moreover, large majorities believed the government’s policies to be ‘too favourable’ to business (n per cent) and ‘too unfavourable’ to workers (n per cent). On the question of what economic policies a future Peronist government should carry out, only n per cent opted for continuity, while n per cent sought to ‘return to the roots of Peronism’. Nevertheless, the PJ’s activist base does not appear to have been substantially eroded in the s. Despite two important elite-level defections (the Group of Eight in  and senator Jose! Octavio Bordo! n in ) and the emergence of the centre-left Front for a Country in Solidarity (FREPASO) as a serious political alternative in the mid-s, "## few PJ activists followed these groups out of the party. Although it is difficult to measure the number of activists who simply quit politics in the s, there is little evidence that the activist base was substantially depleted. Indeed, more than three-quarters of surveyed activists in Greater Buenos Aires said the number of activists either increased ( per cent) or remained the same ( per cent) in the s. Moreover, the relatively high level of PJ activism in the s appears not to have been a product of an influx of new members. Seventy-three per cent of the surveyed activists had worked in the PJ since before , and new activists were at least as likely as older members to oppose Menem’s "#$ policies. Why did anti-Menem activists remain in the party? One reason is patronage. The role of selective material incentives in fostering activist participation increased significantly over the course of the s. More than two-thirds (n per cent) of the UBs surveyed for this study were run by an activist with a government job, and more than a third (n per cent) had two or more activists with government jobs. Moreover, three quarters (n per cent) of the surveyed UBs were financed by agrupaciones with positions in local or provincial governments. As Table  shows, the importance of selective material incentives appears to be increasing over time. Of the surveyed base units that were established before , n per cent were held together primarily by personal ties, social networks, or "#% shared ideology. In only n of the cases were selective material "## In the words of one former Group of Eight activist, ‘no one came with us – not even our wives’ (author’s interview with Mario Wainfeld,  June, ). According to former Group of Eight leader Chacho Alvarez, approximately  activists joined the group (author’s interview,  July, ). "#$ Levitsky, ‘From Laborism to Liberalism’, pp. –. "#% Scorings are the author’s judgments, based on interviews with the activists in the UB. Indicators used were the existence of family relationships, pre-existing friendships, neighbourhood ties, or shared ideology, possession of government jobs or other Local Party Structures in Argentine Peronism 

Table . The increasing role of material benefits in fostering PJ activist participation* (percentage of surveyed base units) UB established UB established between UB established Primary Incentive for before  – after  Activist Participation (n l ) (n l ) (n l ) Personal ties, social networks, nnn or ideology Material benefits nnn * Author’s judgements, based on interviews with activistis in each base unit.  UBs could not be scored due to lack of sufficient information. incentives clearly the dominant linkage between activists and the UB. Of the UBs created between  and , however, the percentage of selective material incentive cases rose to n per cent, and of the UBs that were established after , the percentage rose to n per cent. This evidence suggests that PJ activism is increasingly based on selective, rather than collective, incentives, and that the urban PJ is becoming less "#& of a ‘community of values’ and more of a machine-like party. Yet the stability of the PJ activist base cannot be attributed solely to patronage. Nearly a third of the surveyed UBs (n per cent) had no access to patronage at all, and in a majority of UBs, patronage benefits did extend beyond one or two activists. Hence, even in the late s, a significant number of activists continued to participate despite having little or no access to state resources. Critical to keeping many of these activists in the party was the PJ’s informal structure. Unlike more centralised mass parties such as the Venezuelan AD or the Chilean Communist Party, the PJ’s decentralised structure allowed activists to avoid making a stark choice between adhering to the national party line and leaving (or being expelled from) "#' the party. Specifically, PJ’s system of agrupaciones offered alternative channels of participation to Peronists who disliked the national party’s neoliberal profile. A range of nationalist, traditional populist, social

benefits or stated desire to obtain such benefits, and activist turnover in the base unit. If the activist base in a UB was relatively stable, held few or no government jobs, and showed evidence of social or ideological ties, then the UB was judged to not be based primarily on selective material benefits. "#& Panebianco, Political Parties,p.. "#' Ian McAllister makes a similar argument with respect to the Australian Labor Party (McAllister, ‘Party Adaptation and Factionalism within the Australian Party System’, American Journal of Political Science, vol. , no.  (), pp. –).  Steven Levitsky democratic, and even socialist agrupaciones co-existed with the Menemist national leadership in the s. For example, although the neo-fascist Comando de OrganizacioT n (C de O) abandoned paramilitary activities after , it continued to engage in nationalist activities, such as protests "#( against the British occupation of the Falklands\Malvinas Islands. In , as part of its battle against ‘cultural imperialism’, C de O organised protests against the screening of the (foreign made) film "#) ‘Evita’. "#* An example of a left wing agrupacioT n is ‘March ’ in Quilmes. Founded in  by a group of leftist activists and ex-guerrillas, ‘March ’ grew into the largest agrupacioT n in Quilmes in the s, with approximately  activists. ‘March ’ leaders describe themselves as ‘socialist’ and ‘revolutionary’ and share a commitment to ‘deepening democracy’ through grassroots organisation. They participate in a variety of left-wing political activities, including benefits to raise money for Cuba and an annual party to celebrate the fall of Saigon. Another leftist agrupacioT n is the Federal Capital-based Peronism for Everyone, which maintains a small but committed core of (mainly ex-Montonero) activists that regularly supported strikes and other protests against the Menem government. Finally, scores of agrupaciones – and a much larger number of UBs – provide arenas for participation for what might be called traditional or orthodox Peronists. These activists tend to be strongly attached not only to the traditional Peronist programme, but also to Peronist symbols and practices. An example of a traditional Peronist agrupacioT n is Peronist Loyalty in La Matanza, which is run by former mayor Federico Russo. The second largest agrupacioT n in La Matanza, Peronist Loyalty contains dozens of old guard Orthodox activists, many of whom have worked for Russo since the s. Most of these activists strongly oppose the neo- "$! liberal model, and Russo maintains a populist and anti-liberal profile. Peronist Loyalty engages in a variety of traditional Peronist activities, ‘doctrinal training’, the celebration of Peronist holidays, and the maintenance of women’s and youth branches. "#( The walls of the C de O headquarters in La Matanza are covered with posters that read ‘Defend the Malvinas Islands!’ and ‘Long Live the Argentine Army’. "#) Author’s interview with C de O leader Alberto Brito Lima,  April, . In the Federal Capital, many nationalists belong to Doctrinal Peronism, which is widely believed to have ties to the Carapintada military rebels. "#* This section is based on the author’s interviews with ‘March ’ members Lalo ( May, ), Mario Scalisi ( May, ), Eduardo Schiavo ( April, ), and Oscar Vega ( April, ). "$! In one  assembly, for example, Russo called for the ‘return to a true Justicialista government, made up of Peronists and truly dedicated to national sovereignty and social justice’ (Peronist Loyalty meeting in San Justo, La Matanza,  August, ). Local Party Structures in Argentine Peronism 

Table . Activist responses to the question: ‘‘What level of party activity is most important to you’’?* (in percentages) Federal Capital Greater Buenos Aires Total (n l ) (n l ) (n l ) AgrupacioT n nnn Local\Provincial Party nnn National Party nnn * Based on author’s survey of PJ activists in the Federal Capital, La Matanza, and Quilmes in .

Table . Activists’ views on level of party that is most important, by Ideology (in percentages) Neoliberals Opponents All Activists (n l ) (n l ) (n l ) AgrupacioT n nnn Local\provincial nnn party National party nnn

The PJ’s segmented and decentralised agrupacioT n system thus provided outlets for scores of Peronist activists, allowing them to continue to carry out forms of Peronism that had little to do with – indeed, often contradicted – the programmatic agenda of the Menem government. Data from the activist survey suggests that this base-level autonomy may have induced many activists to stay inside the party. Table  shows activists’ responses to the question, ‘What level of party activity is most important to you?’ In Greater Buenos Aires, nearly two-thirds of the activists said their agrupacioT n (n per cent) or local party (n per cent) was more important to them than the national party. The survey results also suggest that anti-Menem activists were more likely to prioritise their agrupacioT n than were other activists. As Table  shows, nearly half (n per cent) of the activists who characterised themselves as ‘opponents’ of the government’s economic policy viewed their agrupacioT n as the most important level of activity, compared to n per cent of Menem supporters. Taken together, this data suggests is that a substantial number of activists who were critical of the Menem programme took shelter in their agrupaciones in the s, prioritizing local party organisations and, to some extent, detaching themselves from national party activity. In sum, the PJ’s decentralised structure arguably helped the party maintain its activist base in the s. For a relatively small but committed minority, channels existed for the continued expression of leftist, nationalist, or orthodox Peronist beliefs. For a larger group of activists,  Steven Levitsky who were less ideological but nonetheless uncomfortable with the neoliberal turn, the persistence of semi-autonomous local organisations allowed them to continue to practice their own style of Peronism at the grassroots level rather than face a stark choice between Menemism and leaving the party.

Local autonomy and the limits of Menemism Although the persistence of strong, semi-autonomous local organisations provided the PJ leadership with a range of political benefits in the s, it also placed important contraints on that leadership. Because local organisations mediated Menem’s relationship with the Peronist rank-and- file, and because the party bureaucracy lacked the capacity to systematically discipline these sub-units, Menem’s ability to impose strategy or candidates on these sub-units was limited. As a result, Menem was forced to settle for a policy of ‘live and let live’ with those branches. The autonomy of local PJ leaders is in large part rooted in their control over local party machines. Local organisations control the bulk of patronage distribution, mobilise activists, and deliver a large percentage of the party vote. Control over these organisations is essential to winning internal elections. Because voting in primaries is voluntary and virtually all voters must be physically brought to the polling place, winning such elections requires an extensive activist-based organisation. Such organis- ations generally come under the control of local office-holders, such as mayors and governors, who use patronage resources to co-opt agrupaciones into municipal or provincial-level machines. Where such machines consolidate and local bosses gain a monopoly over the local activist base, outsiders – even those backed by the president – stand little chance of success in intra-party competition. Although the national party has the formal authority to intervene provincial branches, doing so in the face of a unified, office-based party is costly, for it risks dividing the party and losing the votes controlled by the local boss. The persistence of strong local machines limited President Menem’s capacity to influence the strategies of lower-level party branches. Although some government officials envisioned a ‘Menemised’ PJ in which neoliberals, business leaders, and pro-Menem ‘outsiders’ would be "$" wedded to the Peronist base through Menem’s direct mass appeal, such

"$" ClarıTn,  Nov., ,p.. Some Menem allies reportedly sought to intervene all non- Menemist provincial branches and impose Menemist leaderships (ClarıTn,  August, ,p.), and others even talked of creating a ‘Menemist party’ that would break with the PJ and base itself on Menem’s direct mass appeal (ClarıTn,  July, ,p.;  Sept., ,p.). Local Party Structures in Argentine Peronism  a transformation did not in fact occur. Efforts to impose strategies on local branches frequently failed, and the national party leadership often found its strategies thwarted – or ignored – by local leaderships. For example, when Menem instructed provincial branches to align with the right wing Centre Democratic Union and other conservative parties in the  elections, only a handful complied. Local leaderships in Salta, San Juan, "$# and other districts openly rejected the order, and many others simply ignored it. Similarly, in , when Menem sought to impose a campaign strategy that centred on the government’s economic programme and "$$ Menem’s re-election, various party branches ignored the national campaign and maintained their own profiles. Indeed, Buenos Aires boss Eduardo Duhalde ordered the ‘de-menemisation’ of the provincial party’s "$% campaign. The Menem leadership was also limited in its capacity to impose candidates on provincial branches. For example, when Menem announced that he planned to support a variety of non-Peronist candidates – including conservative provincial party leaders, ex-military officials, and well-known ‘outsiders’ – who backed his economic programme in the "$&  gubernatorial and legislative elections, provincial branches fiercely resisted the strategy and ultimately forced him to accept party candidacies in all but a few districts. In Buenos Aires, for example, Duhalde ignored Menem’s request to place business leaders Carlos De la Vega and Guillermo Alchouran on the party’s parliamentary list and included only "$' two Menemists in the top  positions on the list. In Mendoza, "$( Menem’s attempt to place ‘people of confidence’ on the legislative list was thwarted when non-Menemist party leaders created their own list and defeated a coalition of Menemist factions in internal elections. Similarly, when government officials designed a strategy in  to ensure that "$) provincial branches nominated Menemist candidates for the senate, "$* they managed – despite months of lobbying – to obtain their preferred candidates in only three districts (the Federal Capital, Entre Rı!os, and Tucuman). In Catamarca, Jujuy, La Pampa, Salta, Santa Cruz, and Santa Fe, the national leadership’s candidates were openly rejected by provincial "$# ClarıTn,  June, ,p.;  July, ,p.. "$$ "$% PaT gina\‚,  March, ,p.. ClarıTn,  Aug., ,p.. "$& ClarıTn,  Nov., ,p.. Menem’s initial list of extra-party gubernatorial candidates included Domingo Cavallo (Co! rdoba), pop singer Palito Ortega (Tucuman), conservative Alberto Natale (Santa Fe), and former military leaders Roberto Ulloa (Salta) and Jose! Ruiz Palacios (Chaco) (ClarıTn,  June, , pp. , ;  June, , p. ;  Oct., ,p.;  May, ,p.). "$' ClarıTn,  June, ,p.;  June, , pp. ;  June, , pp. –;  July, , "$( p. –. ClarıTn,  Oct., ,p.. "$) ClarıTn,  March, ,p.;  March, ,p.. "$* ClarıTn,  May, ,p.;  June, ,p.;  Sept., ,p..  Steven Levitsky "%! party branches. In Santa Fe, for example, despite intense pressure from Menem and other top government officials to re-elect senator Liliana "%" Gurdulich, the local party nominated Jorge Massat, an ally of governor Carlos Reutemann. In La Pampa, where Menem sought the nomination of former governor Nestor Ahuad to fill one of two vacant senate seats, local boss Ruben Marı!n imposed allies Esteban Martinez and Carlos Verna "%# instead. In Buenos Aires, Formosa, Mendoza, Misiones, and San Luis, the national leadership had so little influence that it ultimately decided not to propose a candidate. The Menem leadership also failed to impose its preferred candidate in several key gubernatorial elections. In Mendoza, for example, efforts by top Menemist official Eduardo Bauz to bring the provincial branch ‘fully "%$ in line with the national project of Justicialism’ and nominate non- Peronist businessman Carlos Pulenta for the  gubernatorial candidacy failed when the local party nominated Arturo Lafalla, a Menem critic. In Tucuman, the provincial branch nominated old guard leader Olijela Rivas for the  gubernatorial candidacy despite the public opposition of "%% Menem and intense pressure from government officials. Hence, although Menem was at times able to intervene (or formally replace the leaderships of) provincial party organisations and impose ‘outsider’ Menemist candidates (as in the well-known cases of auto racer Carlos Reutemann and pop singer Palito Ortega in ), these impositions were the exception rather than the rule. In districts in which provincial bosses consolidated stable machines, such as Buenos Aires, Entre Rı!os, Formosa, La Pampa, Mendoza, Misiones, Salta, Santa Cruz, and San Luis, such interventions did not occur. In each of these provinces, party leaders and candidates were consistently selected candidates from within the local organisation, and in almost all of these cases, provincial branches retained traditional ‘Peronist’ – rather than ‘Menemist’ or neo- liberal – profiles. Menem was only able to intervene in provincial branches that were suffering deep internal crises. Such crises occured when corruption scandals discredited sitting governors (as in Santa Fe and Tucuman in ), deep internal conflicts led to the de facto rupture of the party (as in Corrientes, San Juan, Santiago del Estero, and Co! rdoba), or the party – generally out of power – became highly fragmented (as in the Federal Capital). In such cases, important factions sought out the support of the "%! ClarıTn,  June, ,p.;  September, ,p.. "%" El Litoral,  April, ,p.;  May, ,p.;  Sept., ,p.. "%# "%$ ClarıTn,  April, , pp. –. ClarıTn,  March, ,p.. "%% ClarıTn,  Feb., ,p.;  March, ,p.;  March, ,p.. Non- Menemists also won the governorships of Buenos Aires, Entre Rı!os, Salta, Santa Cruz, San Luis, Santiago del Estero, and Santa Fe in . Local Party Structures in Argentine Peronism  national leadership, thereby providing the leadership with the organis- ational base it needed to effectively intervene. Few interventions produced long-term changes in provincial parties, however. In most cases, externally imposed leaderships failed to consolidate control of the party, and in many intervened districts, including Catamarca, Co! rdoba, Santiago del Estero, and Tucuman, old guard leaders soon regained power. In other cases, such as Santa Fe and to a lesser extent San Juan, previously outsider governors built their own support bases within the party and thus gained substantial autonomy from the national leadership. The Menem leadership thus proved relatively limited in its capacity to transform provincial party branches. At the end of Menem’s ten year presidential tenure, the vast majority of provincial branches were governed by non-Menemists and maintained profiles that were far less "%& neo-liberal than that of the national leadership. In many cases, the provincial party was controlled by sectors that had been in power (in some "%' cases, with interruptions) since the early or mid-s. Indeed, only four of the PJ’s  district-level branches – La Rioja, Neuqen, San Juan "%( and the Federal Capital – were controlled by Menemists in . These "%) four districts represented just . per cent of the overall electorate. Contrary to many conventional accounts of the Menem-led PJ, then, traditional Peronism remained largely intact at the local and provincial levels despite the right-wing turn of the national leadership. With few exceptions, the local and provincial organisations that run the PJ’s campaigns, develop its leaders, and select its national legislators became neither Menemist nor neo-liberal. This failure to transform provincial branches helps to explain the rapid erosion of Menem’s influence within the PJ – despite the fact that he remained party president – after he left office in . This outcome highlights the enormous difference between the PJ and other ‘neo-populist’ cases. Whereas Collor’s National Reconstruction Party did not survive the downfall of its leader and "%& Non-Menemist districts included Buenos Aires, Catamarca, Chubut, Co! rdoba, Entre Rı!os, Formosa, La Pampa, Mendoza, Misiones, Salta, San Luis, Santa Cruz, Santiago del Estero, and Tucuman. These districts represent . per cent of the electorate. "%' These include Catamarca, Co! rdoba, Formosa, La Pampa, Mendoza, Misiones, Salta, San Luis, Santiago del Estero, and Tucuman. "%( Of these, only the party leaderships of the Federal Capital and San Juan were neo- liberal. "%) Another indicator of the non-Menemisation of provincial Peronism is the composition of the PJ bloc in congress, for deputies are nominated and elected at the provincial level. In , only  of  PJ deputies belonged to the Menemist sub-bloc, a membership that is smaller than that of the -member Menemist sub-bloc in  (ClarıTn,  June, ,p.). Of the  Menemist deputies, eight were from the Menemist districts of La Rioja and San Juan, and seven more were legislators from the Federal Capital, Co! rdoba, Santa Fe, and Santiago del Estero who were nominated while the party was ‘intervened’ by the national leadership.  Steven Levitsky Fujimori’s various parties will almost certainly face a similar fate, few doubt that the PJ will survive the passing of Menemism.

Conclusion This article has attempted to fill the gap in scholarly work on how the PJ is organised and functions, particularly at the local level. Challenging accounts of the Menem-led PJ as a ‘neo-populist’ party dominated by an unmediated, personalistic leadership, it argues that Peronist leaders and masses have long been linked by a powerful organisational infrastructure with deep roots in urban working and lower class society. The PJ’s mass linkages have been understated and even ignored by scholars because, unlike many European working class parties, they are almost entirely informal. Peronist sub-units organise themselves and maintain only weak ties to the party bureaucracy. Often based out of activists’ homes and rarely registed with party authorities, these sub-units nevertheless constitute a massive base-level infrastructure. This infrastructure yielded the party important political benefits, but it also limited the degree to which President Menem was able to control (or transform) local and provincial party branches. More generally, the Peronist case points to the importance of studying informal patterns of party organisation. Analyses of political parties must go beyond formal structures and examine how parties work in practice. While some parties – for example, many northern European parties – possess relatively formalised or bureaucratic structures, many others, particularly in Latin America, are largely informal. In such cases, studies that focus exclusively on party statutes or formal leadership bodies run the risk of missing the ‘meat’ of the party. This has clearly occured in studies of Peronism, as scholars have often taken the absence of an effective bureaucracy to mean that the party is based primarily on unmediated, personalistic forms of leadership. Not only do such characterisations lack empirical foundation, but they fail to account for the PJ’s capacity to survive – and even thrive – after the passing of its populist (or ‘neo- populist’) leaders. Party founder Juan Pero! n used to say that ‘only organisation conquers time’. Although Pero! n’s party-building never matched his rhetoric, the informal and often chaotic organisation he left behind has proven more resilient – and more effective – than virtually anyone expected. It must therefore be studied more seriously.